RU2754741C1 - Method for adaptation of geological and hydrodynamic reservoir model - Google Patents

Abstract

FIELD: historical data comparison methods.SUBSTANCE: invention generally relates to the comparison of historical data and the prediction of hydrocarbon production from subterranean formations and, in particular, to those methods that use a geological-hydrodynamic model to help optimize the comparison of historical data in order to increase hydrocarbon production. The method includes obtaining, with a given time step, data on the history of wells in the simulated field; obtaining, as a result of the study of the field and wells located on it, the structural and petrophysical parameters of this field. Based on the obtained components and parameters, the values ​​of the dependent parameters for the field are calculated; construction of geological and hydrodynamic models, in which the values ​​of the calculated parameters of the field are changed; calculation of the objective function for each constructed geological and hydrodynamic model, taking into account the calculated and historical values ​​of the components, the time step, normalization to the measurement error, weight coefficients. A geological and hydrodynamic model is selected for field development, which corresponds to the target function with the lowest value. A system and method for adapting the geological and hydrodynamic model of the reservoir, field are also proposed.EFFECT: improvement of comparison of historical data and the prediction of hydrocarbon production from subterranean formations.25 cl, 5 tbl, 5 dwg

Description

The invention generally relates to the comparison of historical data and the prediction of hydrocarbon production from subterranean formations and, in particular, to those methods that use a geological-hydrodynamic model to help optimize the comparison of historical data in order to increase hydrocarbon production.

Known invention under the patent RU 2709047 "Method of adapting the hydrodynamic model of a productive formation of an oil and gas condensate field, taking into account the uncertainty of the geological structure" (priority date: 01/09/2019, publication date: 12/13/2019, IPC: E21B 43/00, E21B 49/00). A method for adapting a hydrodynamic model taking into account the uncertainty of the geological structure of oil and gas condensate fields includes: conducting geophysical, gas-dynamic, hydrodynamic studies of wells, taking a core sample, conducting petrophysical studies, generalizing materials for studying the geological structure, building a geological model of the field, determining the distribution of facies over the area of the field according to petrophysical data. and geophysical studies, determination of effective oil and gas-saturated thicknesses for each cell (block) of a three-dimensional model of the field, assessment of geological indicators in the interwell space, calculation of development indicators on a hydrodynamic model. The development indicators and permissible deviations calculated on the hydrodynamic model are entered into the database of the automatic process control system (APCS) and / or information management system (IMS). They control the actual development indicators for the instruments installed in the wells, and record the results of these measurements in their database. The deviation of the calculated indicators from the actually measured ones is checked. A three-dimensional distribution of the error in calculating the adaptable development indicator is constructed using the hydrodynamic model. An analytical relationship is determined between the geological parameters involved in the adaptation and the adapted development indicator. Determine the values of geological parameters for each cell (block) of the three-dimensional model. For each cell, the correspondence of the values of the given parameters is checked. The adaptation process is continued for other cells (blocks) of the model until the specified error is reached.

This method and adaptation system (separately for each cell of the model) is long, laborious, and the result of the model as a whole may not be comparable with the real regularities of the geological structure of the reservoir.

Known invention under US patent US9703006B2 "Method and system for creating sadapted models" (priority date: 02/12/2010, publication date: 07/11/2017, IPC: G01V11 / 00). A method for correlating production history with a reservoir simulation model includes identifying a plurality of parameters that control an objective function measuring the discrepancy between the simulation result of the parameters in the subspace and the production history. The value of the objective function and the static measurement is calculated in each of the set of experiments in the subspace of parameters. The value is computed for the objective function and for static measurements on each of the many subspace parameter experiments. These results are used to develop a mathematical relationship between one or more static measurements and an objective function. During subsequent adjustments to the model, the objective function is identified and simulations are performed for each modified model, the predictions for which differ from the static geological measurement within the objective function window. The common features of the known technical solution according to the patent US9703006B2 and the claimed methods, system and computer-readable medium is to obtain an adapted geological-hydrodynamic model using the objective function.

Known invention under the patent US8335677B2 "Method of adaptation and quantification of uncertainty using global optimization methods using a proxy" (priority date: 09/01/2006, publication date: 03/27/2008, IPC: E21B43 / 00). The method includes determining the objective function and characteristics of the reservoir model, consistent with the history and taking into account the allowable error. At least one geological model of the formation is created, representing the likely geological setting. For each geological model, a global optimization method is used to perform history matching in a series of iterative steps to obtain acceptable models. The common features of the known technical solution according to the patent US8335677B2 and the claimed methods, system and computer-readable medium is iterative modeling using an objective function.

Known invention under the patent RU 2653251 "Method, device and system for launching the target function" (priority date: 09/23/2015, publication date: 05/07/2018, IPC: H04N 19/00). In the known method, apparatus and system, receiving a preset trigger instruction associated with an objective function is performed; displaying a triggering button of the target function according to a preset triggering instruction; receiving a start instruction generated by triggering an objective function initiation button; and launching the target function according to the launch instruction. The common features of the known technical solution according to the patent RU 2653251 and the claimed system for adapting the geological and hydrodynamic model of the reservoir (field) and the machine-readable medium is the use of instructions implemented by the processor to perform the method of adapting the geological and hydrodynamic model of the reservoir (field).

The existing technical solutions (in methods, systems for adapting geological and geological-hydrodynamic models, as well as machine-readable media created on their basis) have the following disadvantages associated with the lack of accounting for reservoir pressure values in the objective function (CF). In this case, the mutual influence of the CF components is not taken into account and is not compensated for, which can impede effective optimization (with the desired quality and minimum computational costs).

The use of these inventions for the development of deposits (formations) will lead to a decrease in hydrocarbon production.

The ZF formulation is important because its meaning directly affects the optimization process, allowing the modeling process to move in the right direction in finding solutions to increase hydrocarbon production.

The technical result of the invention under consideration is the implementation of a quick and accurate adaptation of the geological and hydrodynamic model of the field formation by determining a more accurate and at the same time simple target function and carrying out on the basis of the selected model of drilling and geological and technical measures in order to increase the volume of hydrocarbon production. Based on the obtained adapted geological and hydrodynamic model (GGDM), the localization of hydrocarbon reserves is carried out, their production, due to which the level of hydrocarbon production increases.

The technical result is achieved due to the fact that the method of adapting the geological and hydrodynamic model of the reservoir (field) includes:

- obtaining, with a given time step, the history data of the wells of the simulated field at least for the following components:

values of flow rates and injectivity of all available types of fluids for wells, reservoir pressures obtained as a result of hydrodynamic studies in wells, and bottomhole pressures for wells;

- obtaining, as a result of the study of the field and wells located on it, the structural and petrophysical parameters of this field;

- on the basis of the obtained components and parameters, the values of the dependent parameters for the field are calculated;

- construction of geological and hydrodynamic models, in which the values of the calculated parameters of the field are changed;

- calculation of the objective function according to the formula for each constructed geological and hydrodynamic model:

где

where

GOF - objective function value;

S - calculated values of the components obtained as a result of calculating the hydrodynamic model;

O - historical values of components;

) либо исторические значения компонентов (О);N - normalization to the measurement error (

) or the historical values of the components (O);

q - component;

i - well;

k is the time step;

n is the number of time steps;

- selection of a geological and hydrodynamic model for field development, which corresponds to the target function with the lowest value.

Also, the technical result is achieved due to the fact that the method for adapting the geological and hydrodynamic model of the reservoir (field) includes:

- obtaining, with a given time step, the history data of the wells of the simulated field at least for the following components:

values of flow rates and injectivity of all available types of fluids for wells, reservoir pressures obtained as a result of hydrodynamic studies in wells, and bottomhole pressures for wells;

- obtaining, as a result of the study of the field and wells located on it, the structural and petrophysical parameters of this field;

- on the basis of the obtained components and parameters, the values of the dependent parameters for the field are calculated;

- construction of geological and hydrodynamic models, in which the values of the calculated parameters of the field are changed;

- determination of two types of weighting coefficients for the objective function:

- weighting factors for wells, which are assigned a value of 1 for wells with reliable measurements or 0 for wells with unreliable measurements,

or weighting factors are assigned in the range from 0 to 1, while the maximum value of the coefficient is assigned for wells located closer to wells with reliable measurements, and the minimum value of the coefficient is assigned for wells located further from wells with reliable measurements,

or weights are assigned in the range from 0 to 1 wells in inverse proportion to the geographic distance of the wells from the zone of interest;

- the weighting factors for the components are calculated from the ratio of the average oil production rate to the value of each component;

- calculation of the objective function according to the formula for each constructed geological and hydrodynamic model:

где

where

GOF - objective function value;

S - calculated values of the components obtained as a result of calculating the hydrodynamic model;

O - historical values of components;

q - component;

i - well;

k is the time step;

) либо исторические значения компонентов (О);N - normalization to the measurement error (

) or the historical values of the components (O);

w _i - weighting factor for wells;

w _q is the weighting factor for the components;

n is the number of time steps;

- selection of a geological and hydrodynamic model for field development, which corresponds to the target function with the lowest value.

Also, the technical result is achieved due to the fact that the method for adapting the geological and hydrodynamic model of the reservoir (field) includes:

- obtaining, with a given time step, the history data of the wells of the simulated field at least for the following components:

values of flow rates and injectivity of all available types of fluids for wells, reservoir pressures obtained as a result of hydrodynamic studies in wells, and bottomhole pressures for wells;

- obtaining, as a result of the study of the field and wells located on it, the structural and petrophysical parameters of this field;

- on the basis of the obtained components and parameters, the values of the dependent parameters for the field are calculated;

- construction of geological and hydrodynamic models, in which the values of the calculated parameters of the field are changed;

- determination of three types of weighting coefficients for the objective function:

- weighting factors for wells, which are assigned a value of 1 for wells with reliable measurements or 0 for wells with unreliable measurements

or weighting factors are assigned in the range from 0 to 1, while the maximum value of the coefficient is assigned for wells located closer to wells with reliable measurements, and the minimum value of the coefficient is assigned for wells located further from wells with reliable measurements,

or weights are assigned in the range from 0 to 1 wells in inverse proportion to the geographic distance of the wells from the zone of interest;

- the weighting factors for the components are calculated from the ratio of the average oil production rate to the value of each component;

- weights for time steps, which are assigned a value of 1 for all time steps,

or a value of 1 is assigned to the time steps of the late adaptation period and 0 to all previous time steps,

- calculation of the objective function according to the formula for each constructed geological and hydrodynamic model:

где

where

GOF - objective function value;

S - calculated values of the components obtained as a result of calculating the hydrodynamic model;

O - historical values of components;

q - component;

i - well;

k is the time step;

σ - measurement error;

w _i - weighting factor for wells;

w _q is the weighting factor for the components;

w _k - weighting factor for time steps;

n - number of time steps

- selection of a geological and hydrodynamic model for field development, which corresponds to the target function with the lowest value.

In one of the options for implementing the methods for adapting the geological and hydrodynamic model of the formation (field), when preparing the history data of wells, in addition, unreliable data of the components are excluded.

In one of the embodiments of the methods for adapting the geological and hydrodynamic model of the reservoir (field), the zone of interest is a section of the field (well), in which adaptation is of priority importance.

In one of the options for implementing the methods for adapting the geological and hydrodynamic model of the reservoir (field), priority components are additionally determined, which are assigned a multiplying coefficient from 1.1 to 3.

In one of the options for implementing the methods for adapting the geological and hydrodynamic model of the reservoir (field), the time steps of the later period are the steps of the period not exceeding the last three years.

In one of the embodiments of the methods for adapting the geological and hydrodynamic model of the reservoir (field), the structural parameters include stratigraphic horizons, reflection coefficients, acoustic impedance, faults and their conductivity obtained from the results of tracer and seismic studies.

In one of the embodiments of the methods for adapting the geological and hydrodynamic model of the formation (field), the petrophysical data include the values of porosity, permeability, saturation, αSP data, G-logging, GG-logging.

In one of the options for the implementation of methods for adapting the geological and hydrodynamic model of the reservoir (field), cellular GGDM are built.

In one of the options for implementing the methods for adapting the geological and hydrodynamic model of the reservoir (field) based on the GGDM, recommendations are additionally prepared for the development of the field.

In one of the embodiments of the methods for constructing geological and hydrodynamic models of the reservoir, in addition, after the selection of the GGDM, the development of the field section is carried out on the basis of the selected GGDM.

The technical result is achieved due to the fact that the adaptation system of the geological and hydrodynamic model of the reservoir (field) includes at least one processor, random access memory and machine-readable instructions for performing any claimed method for adapting the geological and hydrodynamic model.

In one of the embodiments of the adaptation system for the geological and hydrodynamic model of the formation (field), the RAM additionally contains a database with structural, petrophysical parameters and data from the history of wells.

In one of the embodiments of the system for adapting the geological and hydrodynamic model of the reservoir (field), the system additionally contains data allowing access to an external data source.

In one of the embodiments of the system for adapting the geological and hydrodynamic model of the formation (field), the structural and petrophysical parameters and the history of wells are obtained from an external data source.

The technical result is achieved due to the fact that the computer-readable medium contains computer-readable instructions for one of the claimed method of adapting the geological and hydrodynamic model of the reservoir (field), made with the ability to read these instructions and execute them by the processor.

Also, the technical result is achieved due to the fact that the method of field (formation) development includes:

- obtaining a geological and hydrodynamic model (GGDM) as a result of the implementation of a method for adapting a geological and hydrodynamic model of a reservoir (field);

- planning of well drilling and / or geological and technical measures based on the selected GGDM;

- carrying out drilling and / or geological and technical activities based on the selected GGDM;

- production of hydrocarbons.

In one embodiment of the method for developing a field (formation), geological and technical measures include hydraulic fracturing.

In one of the options for implementing the method of developing a field (formation), geological and technical measures include drilling sidetracks.

In one of the options for implementing the method of developing a field (formation), geological and technical measures include treatment of the bottomhole zone.

In one of the embodiments of the method of field (formation) development, when planning drilling, the drilling trajectory and the drilling speed are determined.

The objective function is to estimate the discrepancy between the historical (model values of the previous iteration) and model values of the well operation mode. The objective function reflects the assessment of the quality of adaptation of the obtained geological and hydrodynamic model (GGDM), while the objective function includes at least the values of oil and water production rates, well injectivity, bottomhole pressure and reservoir pressure measurements (determination of the average reservoir pressure). Thus, based on well production data, CF is calculated.

The process of constructing the GGDM and calculating the CF is carried out iteratively. The whole process of constructing the GGDM is described in the form of instructions (workflow) in a specialized software product, for example, Petrel from Schlumberger or tNavigator from RFD. The iterative process can be controlled using an optimization algorithm, for example, an evolutionary strategy. The essence of the optimization algorithm is to change the values of the parameters that determine the model (extreme points for plotting the regression line in functional dependencies, for example, porosity - permeability, values of independent parameters, for example, the conductivity of faults) in order to minimize the values of CF.

To assess the uncertainties of the predicted development indicators, the GGDM with the minimum values of the CF are selected. After the completion of the optimization method / method / algorithm (flattening of the CF, i.e. comparing the values of the CF with the growth of iterations), the GGDM with an acceptable value of the CF, and hence the quality of adaptation, are selected.

The choice of the components included in the target function should be based on the equation of fluid filtration in a porous medium:

- дебит флюида,

- коэффициент продуктивности,

- среднее пластовое давление,

- забойное давление,

- эффективная проницаемость по флюиду,

- толщина коллектора,

- объемный коэффициент флюида,

- коэффициент вязкости флюида,

- радиус контура питания,

- радиус скважины,

- суммарный скин-фактор.where

- fluid flow rate,

- coefficient of productivity,

- average reservoir pressure,

- bottomhole pressure,

- effective fluid permeability,

- collector thickness,

- volumetric coefficient of fluid,

- coefficient of fluid viscosity,

- the radius of the feed loop,

- borehole radius,

is the total skin factor.

Using the example of the well flow rate equation for a steady flow, it is demonstrated which variables vary during adaptation, and which, as a result, should be reproduced by the model. Thus, control is set for historical production rates and wells injectivity. When adapting, it is required to select the values of the parameters that affect the productivity index of the well so that the bottomhole and reservoir pressures are reflected by the model:

On the left side of the equation are the parameters that are found to build a geological and hydrodynamic model, and on the right side of the equation there are historical components, the accuracy of the values of which determines the accuracy and correctness of the adaptation of the geological and hydrodynamic model.

The component composition of the objective function should unambiguously provide historical flow rates, injectivity and pressure drops. The following set of components is recommended:

- current flow rates and injectivity for all types of fluids present;

- bottomhole and reservoir pressures for wells (obtained from hydrodynamic studies of wells - well testing).

The use of current indicators allows avoiding the appearance of accumulated errors, which are inherent in the accumulated indicators of well operation.

Despite the control of the model for fluid, it is necessary to include in the target function the flow rates and injectivity of wells for all types of reservoir fluids, without exception. This approach is justified by the fact that if it is impossible to provide fluid flows at the found reservoir productivity and minimum / maximum bottomhole pressures, fluid control is no longer performed and the model switches to the control mode at a given bottomhole pressure. As a result, when the objective function is minimized, the productivity index will be underestimated. In the case of injection wells, the situation is similar, that is, it is required to include injectivity in the target function.

Since bottomhole and reservoir pressures (according to well testing) are required for correct adaptation and relate to specific wells, it makes sense to use per-well flow rates and injectivity, and not for the field as a whole. In addition, it will allow setting different weighting factors for the wells, if necessary. Numerical modeling on a synthetic model showed the same result in terms of the quality and speed of adaptation in the case of using indicators only for the field, only for wells, and with the simultaneous use of indicators for wells and the field as a whole.

Based on the filtration law, obtaining a reliable model as a result of model adaptation without reservoir pressure data is possible only if the skin factor is constant at all wells and the system is closed, i.e. lack of an active aquifer. Thus, the use of reservoir pressure in the CF makes it possible to most accurately and quickly adjust the geological and hydrodynamic model.

An integral part of calculating the objective function is the historical performance of wells. It is necessary to prepare the following well data, for example, with a monthly time step:

Date (month, year);

Flow rates (oil, gas, water, condensate);

Injectivity (water, gas);

Bottom hole pressure;

Reservoir pressures based on the results of well testing.

Historical data can be loaded in a format appropriate for the simulator used for modeling, for example, into the Petrel simulator. The total and derived parameters can be calculated by the simulator automatically when loading historical data. These include total production rates, average water-cut for the field, water-cut for wells and other indicators.

As a rule, fluid flow rates have a significantly higher measurement accuracy than bottomhole and reservoir pressures. Measurement of oil production is given more attention than water production, as the result is that the oil company needs to sell clean oil.

Normalization to the historical value contributes to the CF in proportion to the relative discrepancy between calculation and measurement. On the one hand, this type of normalization can be simpler and more convenient. The simplicity is that you do not need to specify measurement errors. Convenience is that the dissimilar components of the CF are equally described by one type of discrepancy.

It should be noted that the main goal of specifying the characteristic measurement error is to take into account the accuracy of a particular type of data by the algorithm. If the error is 10 bar and the adjusted pressure differs from the historical one by 10 bar, then this difference lies within the error of this measurement, but the adaptation is considered qualitative. Therefore, it is recommended that measurement uncertainties be specified for objective function components based on physical and engineering understanding, which will allow autoadaptation to be justified. In addition, the risk of getting optimization problems when normalizing to a small value from the history data is eliminated.

When normalizing to measurement errors, there is a risk of getting incommensurate contributions of the components to the objective function in the case when the values of heterogeneous historical data in numerical terms are very different from each other. In such a case, it is suggested to use the historical values of the components.

Determination of the error of historical data can be carried out as follows.

Theoretically, if there were no uncertainties in the distribution of geological and physical properties of the development target, as well as errors inherent in the history data and the method of numerical modeling, then the model would have to reproduce the history of wells with absolute accuracy. However, the inevitability of the named uncertainties and errors leads to the fact that when adapting models, conditional criteria for the quality of adaptation are used, which contain all kinds of errors and express the maximum permissible values of deviations of the results of calculating the model from the corresponding historical indicators.

It is known from the prior art that earlier in the process of adaptation, the hydrodynamic model is "calibrated" on the development history data, which are inherent in certain errors. Measurement error is the deviation of the measurement result from the true value of the measured quantity. In turn, errors give rise to uncertainties in the true values of the measured quantities. In simpler terms, errors in historical data can be divided into three main groups: systematic, random, and gross errors. Systematic errors include errors that remain constant during repeated measurements or change according to some law. The reasons for their appearance include method errors and instrumental errors. It should be understood that in addition to the manufacturer's declared error, there is actual wear and tear of the existing equipment, which can significantly increase the measurement error.

Random error is a component of the measurement error that changes randomly in a series of repeated measurements of the same quantity, carried out under the same conditions. There is no regularity in the appearance of such errors, they are detected during repeated measurements of the same value in the form of a certain scatter of the results obtained. Random errors are inevitable, irreparable and always present as a result of the measurement. The main property of the random error is the ability to reduce the distortion of the desired value by averaging the data.

Gross errors are errors that are not characteristic of the technological process or the result, leading to obvious distortions of the measurement results. Most often, they are admitted by unqualified personnel due to improper handling of the measuring instrument, incorrect readout, recording errors or as a result of a sudden extraneous cause. They are immediately visible among the results obtained, since the obtained values differ from the rest of the values of the set of measurements.

In addition to errors in historical data, there are errors associated with the very process of numerical hydrodynamic modeling. The cause of errors in numerical modeling is the discretization of time and space, that is, the division of space into cells and time into time steps. Numerical variance, grid orientation, averaging of properties during upscaling - all of these factors introduce errors in modeling the operation of reservoirs and wells.

The only type of errors that can be eliminated by improving the results of model adaptation and its subsequent suitability for predictions are gross errors. Thus, before starting the adaptation of the model and calculating the objective function, it is necessary to analyze the accuracy of the historical data, identify and exclude unreliable values from the adaptation process.

When adapting the model, fluid control is usually used. Historical flow rates and liquid injectivity are set in the model, since flow rates are more accurate than pressure data. Bottomhole pressures during autoadaptation should be such as to provide historical flows of formation fluids with a given accuracy.

The weighting factors are used in the two declared options for determining the objective functions to adjust the degree of influence of the constituent quantities on the total value of the objective function. Weights (weighting factors) allow taking into account the different reliability of historical data, balancing strong discrepancies in the numerical expression of the components, and shifting the influence of the components of the objective function in accordance with the priority tasks of adaptation.

Well weights

Identifiers or objects are a reservoir, a group of wells, and individual wells. The weights for the wells are multiplied by the sum of the residuals for the given identifier (wells).

The reasons for using such weights may be different confidence in the quality of measurements for individual wells or the desire to correct the influence of some wells on adaptation in the zone of interest.

In the first case, it is advisable to apply a lowering weight for a well if the quality of measurements of all well performance indicators as a whole is in doubt. For example, well testing was not performed or was of poor quality, there was no downhole pressure sensor, the calculation of the downhole pressure was based on the dynamic level of the gas-liquid mixture in the well annulus, the flow meter had a high degree of wear and gave inaccurate readings. If other wells with more reliable data are involved in the adaptation, it is recommended to assign the weight 0 to the described well. Otherwise, it will be necessary to keep the weight 1 for the well, since there are no other reference points for adaptation (Fig. 1).

Another reason for using weighting factors for wells may be the desire to emphasize the target function on adaptation in the zone of interest. FIG. 2 shows an example of the distribution of weights for adapting the model to predict the performance of a production well planned to be drilled. This approach is most justified under the assumption that the impact of wells is proportional to their distance from the planned drilling point. In the case of a uniform, aged reservoir, this assumption is quite plausible. Wells distant from the zone of interest will contribute less to the target function.

Thus, adaptation will focus on minimizing residuals in the vicinity of the planned drilling.

Component weights

The components of the objective function are the values of the flow rates and injectivity of all the available types of fluid for the wells, the reservoir pressures obtained as a result of the hydrodynamic studies in the wells, and the bottomhole pressures for the wells. The weights for the components are multiplied by the sum of the residuals for that component.

When adapting the model, the values of the partial objective functions that make up the overall global objective function can differ significantly, thereby shifting the focus of adaptation to components with large numerical values. For example, the private target function for bottomhole pressures is expressed in hundreds, for oil flow rates in tens. In this case, in order to minimize the global objective function, the model parameters for iterations will be selected in such a way as to first of all reduce the discrepancies in bottomhole pressures, while minimizing the discrepancies in the oil flow rate fades into the background. As a result, oil production adaptation is likely to be unsatisfactory or will be achieved with a significantly larger number of iterations.

The proposed solution to this problem is balancing the components of the global objective function (objective function for adapting the geological and hydrodynamic model) by weighing the components in inverse proportion to the average oil production rate.

Table 1
Weights for the components of the objective function Objective function component Component weights Oil production rate for all wells 1 Water flow rate for all wells

Water injectivity for all wells

Bottom hole pressure for each well

Reservoir pressure for each well

- средний дебит нефти,

- средний дебит воды,

- средняя приемистость воды,

- забойное давление i-й скважины,

- пластовое давление по ГДИС i-й скважины.

- average oil production rate,

- average flow rate of water,

- average water intake,

- bottomhole pressure of the i-th well,

- reservoir pressure according to well testing of the i-th well.

Thus, each component of the objective function will add a comparable value to the global (final) objective function, allowing the optimizer to adapt the model equally across all components. The effectiveness of this approach is based on analytical conclusions and a series of computational experiments on a synthetic model and a model of a real oil field.

Also, the reason for using weights for components is the different importance of adaptation for different components. For example, if adaptation for gas is more important than for oil, then a multiplying factor is introduced for the gas production rate. Increasing the weight for the priority components will allow you to get models that have been adjusted for these indicators earlier.

Weights for time steps

These weights are multiplied by the residuals at time steps in accordance with the specified distribution.

The assignment of different weights for time steps can be justified in the case when the reliability or accuracy of certain measurements has changed over time.

For example, measurement methods or devices have changed, one subsoil user company with its own standards and norms of operating activities has been replaced by another, and the like.

Also, the reason may be the priority of good adaptation of the last years of history in the case when the forecast of the current trend is important. In this case, at the beginning of the development period, the weighting factor is chosen lower than at the end of the development. When the trend of well performance at the end of the adaptation period is in good agreement with history, there is reason to count on a more reliable forecast of indicators. This approach will help to make more balanced decisions on the development of the field and its sections. In other cases, they are guided by the fact that the indicators on average over the entire adaptation period do not deviate from the permissible values and there is no need to weigh the objective function by time steps.

The method for determining the objective function in the form of the total standard deviation (RMSD) can be determined by the following sequence of steps.

Obtain the results of calculating the model from the hydrodynamic simulator and the corresponding historical indicators that will participate in calculating the given objective function. For example, oil production rates per well (WOPR), bottomhole pressures (WBHP) and others by keywords. Data step, for example, monthly.

Calculate the normalization (residual), for example, on the measurement error between each calculated and historical data point by calculating the difference in values.

Calculate the normalized residual according to one of two predefined options (formulas 1 and 2).

where Δ _k is the normalized residual, S _{q, i, k} is the calculated and O _{q, i, k is the} historical value of the q component (for example, oil production rate) of the identifier i (for example, well P1 or a group of wells) at the time step k, σ is measurement error of the value.

Formula 1 normalizes residuals to measurement errors, formula 2 normalizes to historical values.

The measurement errors for different types of data are different, depending on the types of devices, the measurement principle, the actual wear of the components, on the value of the measured parameters, and others. It is known that the accuracy of measuring the oil production rate is significantly higher than the accuracy of the data on bottomhole pressures in wells calculated through the dynamic level. Normalization to the historical value does not allow setting different levels of errors for the components of the objective function. In addition, computational experiments on a synthetic model have shown that at calculation steps where the historical value is orders of magnitude less than the calculated one, a large contribution to the objective function occurs due to division by a small value. This effect leads to the fact that the optimizer, trying to minimize the value of the objective function, ignores adaptation in other areas where the described effect is absent. With each iteration, the value of the objective function decreases, but the performance indicators of the wells are reproduced worse and worse by the model, that is, such an objective function does not characterize the quality of adaptation.

In connection with the above, it is recommended to use the normalization for the measurement error. In the case of a large difference in the value of the components of the objective function due to the heterogeneity of the data, it is proposed to use balancing weights for the components w _q .

Calculate the total standard deviation for each identifier i and component q using the formula:

where m (i, q) is the total RMSE for each identifier i and component q, w _k are the weight coefficients for time steps k, starting from the moment the well is started (the beginning of the data appearance), n is the number of steps. The weighting factors for time steps can be taken equal to 1, in the range from 0 to 1, or can be excluded from the definition of the CF.

In hydrodynamic simulators, it is technically possible to specify different time weighting factors for each combination of identifier and component.

Further, it is possible to calculate the partial CFs for each component, for example, using the formula:

where POF (q) is the quotient FF (PFC) for the component q, w _q are the weight coefficients for the components, w _i are the weight coefficients for the wells. These weighting factors can be taken equal to 1, in the range from 0 to 1, or can be excluded from the definition of the CF.

As a result, we calculate the global FF (the final FF for the adaptation of the YGDM) using formulas 5 or 6:

where GOF is the objective function or global CF.

As a result, we get three types of target functions, which can be selected depending on the goals of adaptation and initial data:

or

or

In commercial hydrodynamic simulators with the possibility of autoadaptation, the calculation of the objective function is automated. However, the user needs to specify the components, the normalization method and, if necessary, the weighting factors for the objective function, before proceeding to the optimization process.

When implementing the methods, as a result of the study of the field and the wells located on it, the values of the structural and petrophysical parameters of this field are obtained. Structural parameters typically include stratigraphic horizons, reflectivity, acoustic impedance, faults and their conductivities, derived from tracer and seismic surveys. Petrophysical parameters usually include values of porosity, permeability, saturation, incl. capillary pressure curves, logging data (αSP data, G-logging, GG-logging). For example, there may be 150 wells in the field, 20 of which can be used to obtain core data as a result of direct measurements (petrophysical parameters). At the same time, seismic studies can also be carried out at the site, as a result of which data on faults are obtained throughout the entire area of the field. Also, for some wells, well logging (geophysical surveys of wells) can be carried out and petrophysical data obtained. As a result, a set of data is obtained that most closely matches the real situation (have high reliability) and can be accurately used to fill in the model cells, for example, in areas for which there are core data, and also receive data with less accuracy (reliability), which obtained, for example, from the results of GIS.

Further, based on the obtained components and parameters, for example, the values of the dependent parameters for the field are calculated. The calculated data for the YGDM can be obtained based on the functional dependencies between the components and parameters. For example, the dependence of petrophysical parameters is known - the dependence of permeability on porosity, porosity on αSP. Separate functional dependencies can be observed between petrophysical and structural parameters, for example, correlation of values of relative acoustic impedance and porosity, etc.

After that, on the basis of the obtained and calculated values of the parameters and components, the GGDP is obtained, while the GGDM is obtained by slightly changing the values of the calculated parameters of the field (formation).

If the target function is determined according to the first embodiment of the method, then the initial data can be, for example, the following.

WOPR - oil production rate;

WWPR - water flow rate;

WWIR - water injectivity;

WBHP - bottomhole pressure;

WTRP - reservoir pressure.

Legend with "H" at the end - the corresponding indicators for history.

получаем значение целевой функции: 0,98. Substituting the original values into the formula:

we get the value of the objective function: 0.98.

With other calculated data of the components, the value of the CF can be greater or less than 0.98, therefore, the HGMD is selected, for which the CF has a lower value.

When determining the objective function according to the second embodiment of the method with three wells, normalization to historical values and six time steps, weighting coefficients for wells and for components are added. FIG. 3 shows a section of the field with wells located on it.

Well Well weights P1 1.0 P2 0.9 I1 0.7

The weights for the components will be, for example, as follows:

Component Mean Weighting factors for components (oil production / component) Oil flow rate, m ³ / day 144 1.00 Water flow rate, m ³ / day 53 2.73 Water intake, m3 / day 288 0.50 Bottom hole pressure of production wells, atm 70 2.04 Downhole pressure of injection wells, atm 452 0.32 Reservoir pressure of producing wells, atm 180 0.80 Reservoir pressure of injection wells, atm 263 0.55

In this case, we define the target function by the formula:

the resulting objective function value will be 1.87.

Also, the calculated data of the components are changed iteratively, therefore, the CF is calculated iteratively. As a result, the GGDM is selected, which corresponds to the lower value of the CF.

According to the third variant of determining the objective function with three wells, normalization to measurement errors, weighting coefficients for wells and components and six time steps, weighting coefficients for time steps are added.

The weights for the time steps are as follows:

date Weights for time steps 01/01/2020 0 02/01/2020 0 03/01/2020 0 04/01/2020 0 05/01/2020 1 06/01/2020 1

Those. data are excluded from the time period until 05/01/2020.

In this case, we define the target function by the formula:

The resulting objective function value will be 2.12.

As a result of the iterative reconstruction of the GHDM and the calculation of the CF according to the third option, the GGDM is selected, which corresponds to the CF with a lower value .

Also in FIG. 4 clearly shows the degree of influence of the target function on the accuracy of adaptation of the geological and hydrodynamic model, while the declared options for determining the target functions significantly reduce the number of iterations for the adaptation of the geological and hydrodynamic model.

FIG. 5 shows the best HGDM (the obtained distribution of the porosity property is shown) based on the results of adaptation cycles of 200 iterations. Red indicates zones with increased porosity (more than 7%), yellow - medium porosity (4-7%), green - low porosity (less than 4%).

It is possible to compare the CF only when it is determined according to one of the three declared options.

When implementing the method of field development, a GGDM of increased accuracy is obtained, which is determined using the CF according to one of the above options, then well drilling and / or geological and technical measures are planned based on the selected GGDM. Drilling and / or geological and technical measures are carried out on the basis of the selected GGDM and hydrocarbon production.

An example of the implementation of the adaptation system for the geological and hydrodynamic model of the reservoir (field) is similar to the above-described method for adapting the geological and hydrodynamic model of the reservoir (field), while the system includes at least one processor, random access memory and machine-readable instructions for performing the method for adapting the geological and hydrodynamic model of the reservoir ( deposits) according to one of the proposed options.

An example of using the claimed computer-readable medium containing computer-readable instructions for a method for adapting a geological-hydrodynamic model of a reservoir (field) and made with the ability to read these instructions and execute them by the processor is similar to the above sequence of actions.

The proposed methods for adapting the geological and hydrodynamic model of the reservoir (field) were tested on a full-scale geological and hydrodynamic model of an oil field located in Western Siberia. Due to taking into account all aspects: the choice of components, normalization to the error of change and the use of three types of weighting factors, it was possible to achieve a satisfactory quality of adaptation of the model to historical data. After adaptation, a set of selected best models was used to predict oil production in the field over the next 10 years. The result obtained made it possible to take into account potential risks when planning the development of this field.

Claims (91)

1. A method for adapting a geological and hydrodynamic model (GGDM) of a field formation, including: - obtaining, with a given time step, the history data of the wells of the simulated field at least for the following components:
values of flow rates and injectivity of all available types of fluids for wells, reservoir pressures obtained as a result of hydrodynamic studies in wells, and bottomhole pressures for wells; - obtaining, as a result of the study of the field and wells located on it, the structural and petrophysical parameters of this field; - on the basis of the obtained components and parameters, the values of the dependent parameters for the field are calculated; - construction of geological and hydrodynamic models, in which the values of the calculated parameters of the field are changed; - calculation of the objective function according to the formula for each constructed geological and hydrodynamic model:

GOF - objective function value; S - calculated values of the components obtained as a result of calculating the hydrodynamic model; O - historical values of components; N - normalization to the measurement error ( σ ) or the historical values of the components ( O ); q - component; i - well; k is the time step; n is the number of time steps; - selection of a geological and hydrodynamic model for field development, which corresponds to the target function with the lowest value. 2. A method for adapting a geological-hydrodynamic model of a reservoir according to claim 1, in which, when preparing data from the history of wells, in addition, unreliable data of the components are excluded. 3. A method for adapting a geological and hydrodynamic model of a field reservoir, including: - obtaining, with a given time step, the history data of the wells of the simulated field at least for the following components:
values of flow rates and injectivity of all available types of fluids for wells, reservoir pressures obtained as a result of hydrodynamic studies in wells, and bottomhole pressures for wells; - obtaining, as a result of the study of the field and wells located on it, the structural and petrophysical parameters of this field; - on the basis of the obtained components and parameters, the values of the dependent parameters for the field are calculated; - construction of geological and hydrodynamic models, in which the values of the calculated parameters of the field are changed; - determination of two types of weighting coefficients for the objective function: - weighting factors for wells, which are assigned a value of 1 for wells with reliable measurements or 0 for wells with unreliable measurements, or weighting factors are assigned in the range from 0 to 1, while the maximum value of the coefficient is assigned for wells located closer to wells with reliable measurements, and the minimum value of the coefficient is assigned for wells located further from wells with reliable measurements, or weights are assigned in the range from 0 to 1 wells in inverse proportion to the geographic distance of the wells from the zone of interest; - the weighting factors for the components are calculated from the ratio of the average oil production rate to the value of each component; - calculation of the objective function according to the formula for each constructed geological and hydrodynamic model:

GOF - objective function value; S - calculated values of the components obtained as a result of calculating the hydrodynamic model; O - historical values of components; q - component; i - well; k is the time step; N - normalization to the measurement error ( σ ) or the historical values of the components ( O ); w _i - weighting factor for wells; w _q is the weighting factor for the components; n is the number of time steps; - selection of a geological and hydrodynamic model for field development, which corresponds to the target function with the lowest value. 4. A method for adapting a geological and hydrodynamic model of a field formation according to claim 3, in which in the course of preparing the data of the history of wells operation, in addition, unreliable data of the components are excluded. 5. A method for adapting a geological and hydrodynamic model of a field reservoir according to claim 3, in which the area of interest is a field area, where adaptation is of priority importance. 6. A method for adapting a geological and hydrodynamic model of a field reservoir, including: - obtaining, with a given time step, the history data of the wells of the simulated field at least for the following components:
values of flow rates and injectivity of all available types of fluids for wells, reservoir pressures obtained as a result of hydrodynamic studies in wells, and bottomhole pressures for wells; - obtaining, as a result of the study of the field and wells located on it, the structural and petrophysical parameters of this field; - on the basis of the obtained components and parameters, the values of the dependent parameters for the field are calculated; - construction of geological and hydrodynamic models, in which the values of the calculated parameters of the field are changed; - determination of three types of weighting coefficients for the objective function: - weighting factors for wells, which are assigned a value of 1 for wells with reliable measurements or 0 for wells with unreliable measurements, or weighting factors are assigned in the range from 0 to 1, while the maximum value of the coefficient is assigned for wells located closer to wells with reliable measurements, and the minimum value of the coefficient is assigned for wells located further from wells with reliable measurements, or weights are assigned in the range from 0 to 1 wells in inverse proportion to the geographic distance of the wells from the zone of interest; - the weighting factors for the components are calculated from the ratio of the average oil production rate to the value of each component; - weights for time steps, which are assigned a value of 1 for all time steps, or a value of 1 is assigned to the time steps of the late adaptation period and 0 to all previous time steps, - calculation of the objective function according to the formula for each constructed geological and hydrodynamic model:

GOF - objective function value; S - calculated values of the components obtained as a result of calculating the hydrodynamic model; O - historical values of components; q - component; i - well; k is the time step; σ - measurement error; w _i - weighting factor for wells; w _q is the weighting factor for the components; w _k - weighting factor for time steps; n is the number of time steps; - selection of a geological and hydrodynamic model for field development, which corresponds to the target function with the lowest value. 7. A method for adapting a geological and hydrodynamic model of a field formation according to claim 6, in which in the course of preparing the data of the history of wells operation, in addition, unreliable data of the components are excluded. 8. A method for adapting a geological and hydrodynamic model of a field reservoir according to claim 6, in which the area of interest is a field site, where adaptation is of priority importance. 9. The method for adapting the geological and hydrodynamic model of the reservoir formation according to claim 6, in which the priority components are additionally determined, which are assigned a multiplying coefficient from 1.1 to 3. 10. A method for adapting the geological and hydrodynamic model of the field reservoir according to claim 6, in which the time steps of the later period are the steps of a period of no more than the last three years. 11. A method of adapting the geological and hydrodynamic model of the reservoir formation according to any one of paragraphs. 1, 3, 6, in which the structural parameters include stratigraphic horizons, reflection coefficients, acoustic impedance, faults and their conductivities obtained from the results of tracer and seismic studies. 12. The method of adapting the geological and hydrodynamic model of the reservoir reservoir according to any one of paragraphs. 1, 3, 6, in which petrophysical data includes values of porosity, permeability, saturation, αSP data, G-logging, GG-logging. 13. A method for adapting a geological and hydrodynamic model of a reservoir formation according to any one of paragraphs. 1, 3, 6, in which cellular HGDM are built. 14. A method for adapting a geological and hydrodynamic model of a reservoir formation according to any one of paragraphs. 1, 3, 6, in which, on the basis of the GGDM, additional recommendations for the development of the field are prepared. 15. A method for constructing geological and hydrodynamic models of the reservoir formation according to any one of paragraphs. 1, 3, 6, in which, in addition, after the selection of the GGDM, the development of the field is carried out on the basis of the selected GGDM. 16. The system for adapting the geological and hydrodynamic model of the reservoir, including at least one processor, random access memory and machine-readable instructions for performing the method of adapting the geological and hydrodynamic model according to any one of paragraphs. 1-15. 17. The system for adapting the geological and hydrodynamic model of the field reservoir according to claim 16, which additionally contains a database with structural, petrophysical parameters and data from the history of wells in operative memory. 18. The adaptation system of the geological and hydrodynamic model of the field reservoir according to claim 16, which additionally contains data allowing access to an external data source. 19. The adaptation system of the geological and hydrodynamic model of the field reservoir according to claim 16, in which the structural and petrophysical parameters and the data of the well operation history are obtained from an external data source. 20. A computer-readable medium containing computer-readable instructions for a method of adapting a geological and hydrodynamic model of a reservoir formation according to any one of claims. 1-15, configured to read these instructions and execute them by the processor. 21. A method of developing a field formation, including: - obtaining a geological and hydrodynamic model (GGDM) as a result of implementing the method according to any one of paragraphs. 1-15; - planning of well drilling and / or geological and technical measures based on the selected GGDM; - carrying out drilling and / or geological and technical activities based on the selected GGDM; - production of hydrocarbons. 22. The method of developing a field formation according to claim 21, wherein the geological and technical measures include hydraulic fracturing. 23. The method of developing a field formation according to claim 21, wherein the geological and technical measures include drilling sidetracks. 24. The method of field development according to claim 21, in which the geological and technical measures include the treatment of the bottomhole zone. 25. The method of developing a field formation according to claim 21, wherein the drilling trajectory and the drilling speed are determined when planning drilling.

Images